Búsqueda Imágenes Maps Play YouTube Noticias Gmail Drive Más »
Iniciar sesión
Usuarios de lectores de pantalla: deben hacer clic en este enlace para utilizar el modo de accesibilidad. Este modo tiene las mismas funciones esenciales pero funciona mejor con el lector.

Patentes

  1. Búsqueda avanzada de patentes
Número de publicaciónUS7536637 B1
Tipo de publicaciónConcesión
Número de solicitudUS 12/027,659
Fecha de publicación19 May 2009
Fecha de presentación7 Feb 2008
Fecha de prioridad7 Feb 2008
TarifaCaducada
Número de publicación027659, 12027659, US 7536637 B1, US 7536637B1, US-B1-7536637, US7536637 B1, US7536637B1
InventoresAndreas Nauerz, Stefan Liesche, Stefan Behl, Michael Junginger
Cesionario originalInternational Business Machines Corporation
Exportar citaBiBTeX, EndNote, RefMan
Enlaces externos: USPTO, Cesión de USPTO, Espacenet
Method and system for the utilization of collaborative and social tagging for adaptation in web portals
US 7536637 B1
Resumen
This application explores the value of Web 2.0 techniques to perform adaptation based on semantic annotations. Semantic annotations are used to derive user and context models which in turn are used for performing the re-ordering of page layouts or the dynamic transformation of navigation topologies to provide a more user specific portal user interface. The semantic annotations also allow similarities and relationships to be calculated between resources and users so that additional relevant annotations and/or resources can be recommend to the user.
Imágenes(4)
Previous page
Next page
Reclamaciones(1)
1. A computer implemented method of providing a navigation topology in a portal supporting a social network, said method comprising:
annotating portal resources;
wherein said portal resources are uniquely identifiable fragments;
wherein said portal resources comprise one or more of the following: page, portlet, user, email, wiki, blog post, or applet;
wherein said annotation is performed by a user, or is performed automatically by extracting some data from first portal resources among said portal resources visited by said user;
an automated bot annotating said portal resources;
storing said user generated annotations and said bot generated annotations in an annotation store accessible by a plurality of different annotators;
tracking said user's annotating behavior, based on all annotations applied to said portal resources which said user visits or uses;
deriving a user model based on all users' annotating behavior, said all users' preference and interest, and said all users' profiles, using a web mining method;
wherein said web mining method comprises at least text mining;
assigning said user to a group model based on said web mining method and recommending changes made by a specific fraction of other group members of said group model to said user;
deriving a context model based on one or more of factors;
wherein said one or more of factors comprise one of said user's annotating behavior, timing of resource annotations, a device said user is working with, and a location of said user;
determining similarities and relationships between said portal resources and tags, by calculating a semantic distance between said tags and said portal resources;
partitioning said user model into a sole partition for each context profile available in said context model;
automatically determining a matching profile, by observing a set of predefined context attributes;
wherein said user has an option to override said matching profile;
dynamically adopting a navigation topology for each of said all users, by generating a tree model at run time, based on behavior of said each of said all users and current context;
wherein a tag-driven structure has a minimum click distance between content elements that are annotated;
wherein said navigation topology comprises a dynamic node;
wherein said dynamic node is associated with a transformation;
wherein said dynamically adopting said navigation topology comprises constructing a topology by traversing said tree model, and invoking said transformation;
wherein said transformation performs its specific operation to generate a sub-model;
wherein said specific operation comprises one of moving or hiding navigation tree nodes depending on their relevance;
wherein said sub-model is phased into said navigation topology, directly underneath said dynamic node; and
displaying said constructed topology to said user.
Descripción
RELATED APPLICATION

This application is related to an application titled “Method and System for Construction of Social Networks and Expert Determination”, U.S. Ser. No. 12/027,686, filed on Feb. 7, 2008.

BACKGROUND OF THE INVENTION

Different situations lead to different requirements. Within Web Portal environments the requirements influenced might affect the entire presentation layer, especially with entities such as the navigation or page layout. Currently, the adaptation is based on what central instances like administrators do based on observation of single set of users or user groups. Within the context of Web 2.0 paradigm, contents are now no longer centrally controlled since they comprise of too many pages generated by users and communities that current strategies and algorithms for adaptation neglect. Current models and algorithm for adaptation to different requirements neglect the specificity of each user.

SUMMARY OF THE INVENTION

An embodiment of this invention explores the value of Web 2.0 techniques to further enhance the strategies to perform adaptation, based on semantic annotations.

With semantic annotation, we denote the act of “tagging”—the association of words or phrases with a resource. Tagging functionalities have been built into several web application introduced in recent years (Flickr®, delicious, dogear, citeulike).

An embodiment of the invention uses semantic annotations (i.e. adding semantics to resources) and provides navigation through the semantic space via tag clouds. It performs automatic extraction of data for analysis. Analysis of unstructured data involves a process of deriving high quality information from text. For example, high quality information is derived through the learning of patterns and trends through means such as statistical pattern learning. The embodiment takes advantage of text mining tools which involves the process of structuring the input text (usually parsing, along with the addition of some derived linguistic features and the removal of others, and subsequent insertion into a database), deriving patterns within the structured data, and finally evaluation and interpretation of the output. The quality of text mining depends on some combination of relevance, novelty, and interestingness. Typical text mining tasks include text categorization, text clustering, concept/entity extraction, production of granular taxonomies, sentiment analysis, document summarization, and entity relation modeling (i.e., learning relations between named entities). Accordingly, an embodiment of this invention automatically constructs tags describing the semantics behind objects to perform adaptation, and provides a dynamic navigational topology which based on current context of the user.

BRIEF DESCRIPTION OF THE DRAWINGS

FIG. 1 is a schematic diagram of the tagging and adaptation layer framework

FIG. 2 is a schematic diagram of the modeling and annotation layers

FIG. 3 is the flow diagram of tagging and annotation

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

While users are applying annotations we assume they are expressing interest to or knowledge about some topics. Despite some annotations which are only useful to the user herself (e.g. “to read” a private annotation), annotations add some valuable insight into the semantics of the resource and hence into the interests of the user.

The user model in one embodiment is further refined when the annotations applied to this user as a resource. Matching annotations will highly likely indicate an expert in a specific field. By using this kind of annotations a user model is also built for users who do not use the annotation features at all.

In one embodiment, other knowledge about the users' behavior or interests is deduced into (user) group models which subsume the similarities of some users.

In an embodiment, knowing the users' interests or expert knowledge helps generate dynamic buddy lists showing like minded or specialist users related to some resource(s) a user currently works with.

On embodiment assigns the user to a group to recommend the latest changes made by some group members to the others (e.g. a portlet lately annotated as helpful by some users will be recommended to the whole group).

Context Model

One embodiment constructs the context model based on several knowledge sources, e.g., (1) all annotations applied to the resource artifacts a user is currently working with, (2) the creation date and time of resource related annotations, (3) the device a user is working with, (4) the location of the user.

In one embodiment, different context scenarios each form a context profile. E.g., the “traveling” profile may apply to a user using a mobile device while she is not in the office, in contrast to her associate who is present in the office. While in one context profile, a user may uses a very specific vocabulary for annotations and the same may apply to the resources she is working with. As example, a traveling user noticeably often uses resources annotated with “mobile”, “schedule”, “map”, and also, while traveling, she tends to apply these annotations to resources herself.

Semantic-Based Adaptation

One embodiment uses semantic-based adaptation to perform link-level and content-level adaptation. It re-orders the resources which are of interest for the user on the page, or it sorts or hides links to resources to facilitate personalization. By utilizing the derived user and context models different similarity calculations can be performed in an embodiment of the invention.

Examples of methods for determining such similarity values include (1) similarity calculation based on the number of common annotation (and their synonyms) and (2) cosine similarity between two concepts using Latent Semantic Indexing (LSI). The first one will give a higher similarity value if two resources share more annotations or as a variant two annotations will be treated as more similar if they often appear together.

In one embodiment, computing the semantic distance between two resources annotated by a user (e.g. based on the number of common annotation) or between the annotations of the resources themselves will provide a value indicating the relatedness of these resources.

One embodiment performs re-ordering of the pages layout or transformation of the navigation topology and provides a more user specific portal UI. For example, based on the users' interests the more important portlets on a page are grouped at the beginning; or while performing a specific task, portlets marked as helpful are shown dynamically and unnecessary content vanishes. One embodiment moves or hides navigation tree nodes depending of their relevance.

Performing the semantic distance calculation on a resource and a user in one embodiment shows of how much interest the resource is for this user. When using this metric to compare two users, it indicates how related their interests are. To utilize the results of these comparisons, an embodiment provides recommendation (330) to advice other resources (e.g. pages, portlets, expert users) which are highly likely of interest in the current context.

One embodiment performs adaptation in the user group analysis. If a specific fraction of group members lately applied the same annotation(s) to a new resource, the embodiment recommends this resource to the other users, as well.

In one embodiment, an architecture allows any kind of annotator be it a human user or a programmatic one to annotate any resource within web portals (FIG. 3). The data will then be used by modeling and adaptation components that facilitate different adaptation methods. One embodiment provides tailor-made navigation structures and page layout as well as recommendations (330) and suggestions for each individual user to interact with (332) based on automatically generated semantic annotations (322).

One embodiment provides a generic annotation store (220) which can be accessed by different kinds of annotators (212) via an annotation layer (218). This will enable user generated annotations to be stored along with automatically extracted semantics, e.g. by specialized annotators running in a framework like Unstructured Information Management Architecture (UIMA). Tagging and adaptation framework (110) for a web portal (112) comprises an adoption layer (used to generate navigation topology for a given user in a given context within the web portal), modeling layer (which manages both user model and context model) (210), and annotation layer (which provides an interface to annotators (including users) to tag resources of the portal and support creation and management of tags).

One embodiment of the invention is a method of providing a navigation topology in a portal (112 or 310) supporting a social network, the method comprising:

    • annotating portal resources; wherein the portal resources are uniquely identifiable fragments; wherein the portal resources comprise one or more of the following: page, portlet (216), user, email, wiki, blog post, or applet (310); wherein the annotation is performed by a user (312), or is performed automatically by extracting some data from first portal resources among the portal resources visited by the user;
    • an automated bot annotating the portal resources;
    • tracking the user's annotating behavior (314), based on all annotations applied to the portal resources which the user visits or uses;
    • deriving a user model (210 or 316) based on all users' annotating behavior, preference and interest, and profiles, using a web mining method;
    • deriving a context model (316) based on one or more of factors; wherein the one or more of factors comprise of the user's annotating behavior, timing of resource annotations, a device the user is working with, and a location of the user;
    • determining similarities and relationships between the portal resources (320) and tags (318), by calculating a semantic distance between the tags and the portal resources;
    • partitioning the user model into a sole partition for each context profile available in the context model;
    • automatically determining a matching profile, by observing a set of predefined context attributes; wherein the user has an option to override the matching profile;
    • when users select tags (326) and resources (328), recommending related tags (324) and related resources (330).
    • dynamically adopting a navigation topology for each of the all users, by generating a tree model at run time, based on behavior of the each of the all users and current context. The method uses a tag-driven structure (322) with a minimum click distance between content elements that are annotated. The navigation topology comprises of a dynamic node; and the dynamic node is associated with a transformation. To dynamically adopt the navigation topology, the method constructs topology by traversing the tree model, and invoking the transformation. The transformation performs its specific operation to generate a sub-model; and the sub-model is phased into the navigation topology, directly underneath the dynamic node.

A system, apparatus, or device comprising one of the following items is an example of the invention: navigation topology, portal, social network, portal resources, annotation, page, portlet, user, email, wiki, blog post, applet, automated bot, user model, user preference and interest, user profile, web mining tool, context model, tag, semantic distance, tree model, click distance, node, dynamic node, sub-model, or any software, applying the method mentioned above, for purpose of invitation or providing a navigation topology in a portal supporting a social network.

Any variations of the above teaching are also intended to be covered by this patent application.

Citas de patentes
Patente citada Fecha de presentación Fecha de publicación Solicitante Título
US6327628 *19 May 20004 Dic 2001Epicentric, Inc.Portal server that provides a customizable user Interface for access to computer networks
US6636242 *31 Ago 199921 Oct 2003Accenture LlpView configurer in a presentation services patterns environment
US6832263 *27 Abr 200114 Dic 2004Hyperion Solutions CorporationMethod and apparatus for implementing a dynamically updated portal page in an enterprise-wide computer system
US6968333 *2 Abr 200122 Nov 2005Tangis CorporationSoliciting information based on a computer user's context
US7155678 *31 Ago 200126 Dic 2006Gxs, Inc.Method, system, and software for generating and displaying custom views for accessing multiple applications
US72750865 Ago 199925 Sep 2007Intellisync CorporationSystem and method for embedding a context-sensitive web portal in a computer application
US7433876 *19 Feb 20057 Oct 2008Radar Networks, Inc.Semantic web portal and platform
US7457814 *30 May 200625 Nov 2008Microsoft CorporationMulti-dimensional data classification for user interface customizations
US20050256866 *15 Mar 200517 Nov 2005Yahoo! Inc.Search system and methods with integration of user annotations from a trust network
US20050256906 *13 May 200517 Nov 2005Bea Systems, Inc.Interface for portal and webserver administration-efficient updates
US20060004680 *11 Ene 20055 Ene 2006Robarts James OContextual responses based on automated learning techniques
US20060036993 *26 Jul 200516 Feb 2006International Business Machines CorporationSystem and method for developing portal applications and for automatically deploying portal applications into a portal server application
US20060112146 *31 Mar 200525 May 2006Nec Laboratories America, Inc.Systems and methods for data analysis and/or knowledge management
US200602828199 Jun 200514 Dic 2006International Business Machines CorporationGeneral purpose annotation service for portal-based applications
US20060294086 *2 Ago 200628 Dic 2006Yahoo! Inc.Realtime indexing and search in large, rapidly changing document collections
US20070038610 *15 Ago 200615 Feb 2007Nosa OmoiguiSystem and method for knowledge retrieval, management, delivery and presentation
US200701131949 Mar 200617 May 2007Bales Christopher ESystem and method for providing drag and drop functionality in a communities framework
US200701566363 Ene 20065 Jul 2007Yahoo! Inc.Apparatus and method for controlling content access based on shared annotations for annotated users in a folksonomy scheme
US2007020400417 Nov 200630 Ago 2007Qualcomm IncorporatedApparatus and methods of distributing content and receiving selected content based on user personalization information
US200702260773 Mar 200627 Sep 2007Frank Martin RCollaborative Structured Tagging for Item Encyclopedias
US20080040301 *10 Ago 200614 Feb 2008Yahoo! Inc.System and method for inferring user interest based on analysis of user-generated metadata
US20080052372 *19 Ene 200728 Feb 2008Yahoo! Inc.Method and system for presenting information with multiple views
US20080066002 *8 May 200713 Mar 2008Andreas NauerzMethod for creation, management, and presentation of user-scoped navigation topologies for web applications
US20080086458 *14 Sep 200710 Abr 2008Icebreaker, Inc.Social interaction tagging
US20080092044 *12 Oct 200617 Abr 2008International Business Machines CorporationCascading clouds
US20080189265 *6 Feb 20077 Ago 2008Microsoft CorporationTechniques to manage vocabulary terms for a taxonomy system
US20080195664 *13 Dic 200714 Ago 2008Quickplay Media Inc.Automated Content Tag Processing for Mobile Media
Otras citas
Referencia
1 *Smyth et al, "Relevance at a Distance-An Investigation of Distance-Baised Personalization on the Mobile Internet", 2003, pp. 1-6, http://www.csi.ucd.ie/UserFiles/publications/1125325052899.pdf.
Citada por
Patente citante Fecha de presentación Fecha de publicación Solicitante Título
US7676557 *16 Ene 20099 Mar 2010International Business Machines CorporationDynamically adaptive portlet palette having user/context customized and auto-populated content
US8060451 *15 Jun 200715 Nov 2011International Business Machines CorporationSystem and method for facilitating skill gap analysis and remediation based on tag analytics
US8166069 *29 Mar 200624 Abr 2012Yahoo! Inc.Displaying aggregated new content by selected other user based on their authorization level
US8255493 *10 Dic 200928 Ago 2012International Business Machines CorporationAutomated dynamic portlet rendering
US82808921 Oct 20082 Oct 2012Fujitsu LimitedSelecting tags for a document by analyzing paragraphs of the document
US838071613 May 201019 Feb 2013Jan MirusMind map with data feed linkage and social network interaction
US849504826 Ago 200923 Jul 2013International Business MachinesApplying user-generated deployment events to a grouping of deployable portlets
US8495081 *14 Dic 201023 Jul 2013International Business Machines CorporationMethod, system and computer program product for federating tags across multiple systems
US8583658 *22 Jun 201112 Nov 2013International Business Machines CorporationCreating and managing reference elements of deployable web archive files
US858425820 Dic 200512 Nov 2013Yahoo! Inc.Control for inviting an unauthenticated user to gain access to display of content that is otherwise accessible with an authentication mechanism
US8612209 *20 Sep 201117 Dic 2013Nuance Communications, Inc.Classifying text via topical analysis, for applications to speech recognition
US861235914 Jun 200417 Dic 2013Yahoo! Inc.Method and system for sharing portal subscriber information in an online social network
US86157114 Ene 201124 Dic 2013International Business Machines CorporationSummarizing portlet usage captured responsive to trigger events in a portal page
US878895314 Mar 201222 Jul 2014International Business Machines CorporationSummarizing portlet usage in a portal page
US20100306249 *27 May 20102 Dic 2010James HillSocial network systems and methods
US20100325557 *22 Jun 201023 Dic 2010Agostino SibilloAnnotation of aggregated content, systems and methods
US20110145364 *10 Dic 200916 Jun 2011International Business Machines CorporationAutomated dynamic portlet rendering
US20110271171 *22 Jun 20113 Nov 2011International Business Machines CorporationCreating and Managing Reference Elements of Deployable Web Archive Files
US20110314014 *14 Dic 201022 Dic 2011International Business Machines CorporationMethod, system and computer program product for federating tags across multiple systems
US20120010875 *20 Sep 201112 Ene 2012Nuance Communications Austria GmbhClassifying text via topical analysis, for applications to speech recognition
US20120192082 *13 Ene 201226 Jul 2012International Business Machines CorporationPersonalization of web content
US20120284605 *6 May 20118 Nov 2012David H. SitrickSystems And Methodologies Providing For Collaboration Among A Plurality Of Users At A Plurality Of Computing Appliances
EP2463788A1 *6 Dic 201113 Jun 2012Alcatel LucentSystem for selecting key words, method and database implementing such a system
Clasificaciones
Clasificación de EE.UU.715/230, 715/853, 715/234, 704/9, 715/744, 715/247, 706/55, 707/999.107, 707/999.104
Clasificación internacionalG06F17/30, G06F15/18, G06F17/21
Clasificación cooperativaG06F17/3089, Y10S707/99945, Y10S707/99948
Clasificación europeaG06F17/30W7
Eventos legales
FechaCódigoEventoDescripción
9 Jul 2013FPExpired due to failure to pay maintenance fee
Effective date: 20130519
19 May 2013LAPSLapse for failure to pay maintenance fees
31 Dic 2012REMIMaintenance fee reminder mailed
27 Feb 2008ASAssignment
Owner name: INTERNATIONAL BUSINESS MACHINES CORPORATION, NEW Y
Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:NAUERZ, ANDREAS;LIESCHE, STEFAN;BEH, STEFAN;AND OTHERS;REEL/FRAME:020570/0507;SIGNING DATES FROM 20080129 TO 20080131